• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권2호
• /
• pp.522-539
• /
• 2021

In order to solve the problems of the existing audio fingerprint method when extracting audio fingerprints from long speech segments, such as too large fingerprint dimension, poor robustness, and low retrieval accuracy and efficiency, a robust audio fingerprint retrieval method based on feature dimension reduction and feature combination is proposed. Firstly, the Mel-frequency cepstral coefficient (MFCC) and linear prediction cepstrum coefficient (LPCC) of the original speech are extracted respectively, and the MFCC feature matrix and LPCC feature matrix are combined. Secondly, the feature dimension reduction method based on information entropy is used for column dimension reduction, and the feature matrix after dimension reduction is used for row dimension reduction based on energy feature dimension reduction method. Finally, the audio fingerprint is constructed by using the feature combination matrix after dimension reduction. When speech's user retrieval, the normalized Hamming distance algorithm is used for matching retrieval. Experiment results show that the proposed method has smaller audio fingerprint dimension and better robustness for long speech segments, and has higher retrieval efficiency while maintaining a higher recall rate and precision rate.

• Journal of information and communication convergence engineering
• /
• 제1권1호
• /
• pp.53-57
• /
• 2003

Chaos theory is a study researching the irregular, unpredictable behavior of deterministic and non-linear dynamical system. The interpretation using Chaos makes us evaluate characteristic existing in status space of system by tine series, so that the extraction of Chaos characteristic understanding and those characteristics enables us to do high precision interpretation. Therefore, This paper propose the new method which is adopted in extracting character features and recognizing characters using the Chaos Theory. Firstly, it gets features of mesh feature, projection feature and cross distance feature from input character images. And their feature is converted into time series data. Then using the modified Henon system suggested in this paper, it gets last features of character image after calculating Box-counting dimension, Natural Measure, information bit and information dimension which are meant fractal dimension. Finally, character recognition is performed by statistically finding out the each information bit showing the minimum difference against the normalized pattern database. An experimental result shows 99% character classification rates for 2,350 Korean characters (Hangul) using proposed method in this paper.

• 한국CDE학회논문집
• /
• 제1권2호
• /
• pp.122-132
• /
• 1996

Features are generic shapes with which engineers associate certain attributes and knowledge useful in reasoning about the product. Feature-based modeling systems support additional levels of information beyond those available in geometric modelers. The objective of this study is to develop a PC level feature-based modeling system which explicitly represents dimensions of the part. The feature-based modeler retains all the benefits of traditional B-rep. solid models, and represents the dimensions at a high level of a abstraction so that dimension driven geometry can be achieved.

• 정보처리학회논문지B
• /
• 제15B권4호
• /
• pp.285-292
• /
• 2008

색상 히스토그램은 영상의 색상 특징을 표현하기 위한 특징 벡터로 빈번히 사용되지만, 고차원의 특징 벡터를 생성하므로 효율성의 면에서 한계점을 갖고 있다. 본 논문에서는 주어진 차량 영상의 색상 히스토그램에 PCA (principal components analysis) 기법을 적용하여 특징 벡터의 차원을 축소시키는 방법을 제안한다. 차원이 축소된 특징 벡터들에 대해서는 SVM (support vector machine) 기법을 적용하여 차량 색상을 인식하기 위해 사용한다. 특징 벡터의 차원을 1/32로 축소한 결과, 차원이 축소되기 이전의 특징 벡터와 비교하여 약 1.42%의 미소한 차이로 색상 인식 성공률이 감소하였다. 또한, 색상 인식의 수행 시간은 1/31로 단축됨으로써 효율적으로 색상 인식을 수행할 수 있었다.

• 융합신호처리학회논문지
• /
• 제7권3호
• /
• pp.116-121
• /
• 2006

본 논문에서는 다양한 잡음환경에서의 실시간 적응화 기법을 적용하기 위한 선결 과제로 다차원 음성 특정 벡터를 저차원으로 축소하는 방법을 제안한다. 제안된 방법은 특징 벡터를 확률 우도 값으로 매핑시켜 비선형적으로 축소하는 방법으로 음성 / 비음성의 분류는 우도비 검증 (Likelihood Ratio Test; LRT) 을 이용하여 분류하였다. 실험 결과 고차원 특징 벡터를 이용하여 분류한 결과와 대등하게 분류됨을 확인할 수 있었다. 그리고, 제안된 방법에 의해 검출된 음성 데이터를 이용한 음성인식 실험에서도 10차 MFCC(Mel-Frequency Cepstral Coefficient)를 사용하여 분류한 경우와 대등한 인식률을 보여주었다.

• Journal of Electrical Engineering and Technology
• /
• 제13권5호
• /
• pp.2099-2106
• /
• 2018

This study presents a novel approach of discriminative feature vectors based on manifold learning using nonlinear dimension reduction (DR) technique to improve loss function, and combine with the Adversarial examples to regularize the object function for image classification. The traditional convolutional neural networks (CNN) with many new regularization approach has been successfully used for image classification tasks, and it achieved good results, hence it costs a lot of Calculated spacing and timing. Significantly, distrinct from traditional CNN, we discriminate the feature vectors for objects without empirically-tuned parameter, these Discriminative features intend to remain the lower-dimensional relationship corresponding high-dimension manifold after projecting the image feature vectors from high-dimension to lower-dimension, and we optimize the constrains of the preserving local features based on manifold, which narrow the mapped feature information from the same class and push different class away. Using Adversarial examples, improved loss function with additional regularization term intends to boost the Robustness and generalization of neural network. experimental results indicate that the approach based on discriminative feature of manifold learning is not only valid, but also more efficient in image classification tasks. Furthermore, the proposed approach achieves competitive classification performances for three benchmark datasets : MNIST, CIFAR-10, SVHN.

• 한국음향학회지
• /
• 제36권2호
• /
• pp.144-150
• /
• 2017

음악 장르는 음악 검색 및 분류 등의 정보 처리 시스템 구현에 있어서 필수적인 요소이다. 일반적으로 장르 분류를 위한 스펙트럼 특징은 음악의 화음 및 강약 구조를 표현하기 위해 부밴드로 분해하여 구해진다. 본 논문은 음악 장르 분류 성능 개선을 위한 특징 추출을 위한 부밴드 분해 방법에 관해 연구하였다. 또한 부밴드 음악 특징의 차수를 줄일 수 있는 방법에 대해서도 연구하였다. 널리 사용되고 있는 장르 데이터셋들에서 실험을 수행하여 널리 사용되고 있는 옥타브 스케일보다 세분화된 부밴드 분해가 장르 분류 성능을 향상시킬 수 있으며, 특징 차수 축소를 결합하여 분류기의 계산량도 줄일 수 있음을 보였다.

• 한국정밀공학회지
• /
• 제15권6호
• /
• pp.167-174
• /
• 1998

The ultrasonic test is recognized for its significance as a non-destructive testing method to detect volume defects such as porosity and incomplete penetration which reduce strength in the weld zone. This paper illustrates the defect detection in the weld zone of ferritic carbon steel using ultrasonic wave and the evaluation of pattern recognition by chaotic feature extraction using time series signal of detected defects as data. Shown in the time series data were that the time delay was 4 and the embedding dimension was 6 which indicate the geometric dimension of the subject system and the extent of information correlation. Based on fractal dimension and lyapunov exponent in quantitative chaotic feature extraction, feature value of 2.15, 0.47 is presented for porosity and 2.24, 0.51 for incomplete penetration The precision rate of the pattern recognition is enhanced with these values on the total waveform of defect signal in the weld zone. Therefore, we think that the ultrasonic pattern recognition method of weld zone defects of ferritic carbon steel by ultrasonic-chaotic feature extraction proposed in this paper can boost precision rate further than the existing method applying only partial waveform.

• 한국멀티미디어학회논문지
• /
• 제24권12호
• /
• pp.1614-1623
• /
• 2021

In this paper, the purpose is evaluation of the effect of using fractal feature in machine learning based pancreatic tumor classification. We used the data that Pancreas CT series 469 case including 1995 slice of benign and 1772 slice of malignant. Feature selection is implemented from 109 feature to 7 feature by Lasso regularization. In Fractal feature, fractal dimension is obtained by box-counting method, and hurst coefficient is calculated range data of pixel value in ROI. As a result, there were significant differences in both benign and malignancies tumor. Additionally, we compared the classification performance between model without fractal feature and model with fractal feature by using support vector machine. The train model with fractal feature showed statistically significant performance in comparison with train model without fractal feature.

• 정보처리학회논문지B
• /
• 제15B권5호
• /
• pp.435-448
• /
• 2008

사용자 포즈의 3차원 데이터 생성을 통한 3차원 포즈 인식은 2차원 포즈 인식의 문제점을 해결하기 위해서 많이 연구되고 있지만, 3차원 표면 데이터의 방대한 양으로 포즈 인식에서 중요한 특징 추출(feature extraction)이 어렵고 수행 시간이 많이 걸리는 문제점을 가지고 있다. 본 논문에서는 3차원 포즈 인식의 두 가지 문제점인 특징 추출의 어려움과 느린 처리속도를 개선하기 위해서 3차원 형상복원 기술로 모델의 3차원 표면 점들로 구성된 데이터를 2차원 데이터로 변환하는 차원 축소(dimension reduction) 방법을 제안한다. 실린더형 외곽점을 이용한 메쉬없는 매개변수화(meshless parameterization) 방법은 방대한 데이터인 3차원 포즈 데이터를 2차원 데이터로 변환하여 특징 추출과 매칭과정의 연산 속도를 향상 시키며, 특징 추출의 효율성 검증을 위해 간단한 환경에서 실험이 가능한 손 포즈 인식 및 인간 포즈 인식에 적용하였다.